Article(id=1243896551068775110, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1243896547704947617, articleNumber=null, orderNo=null, doi=10.3969/j.issn.1007-7294.2025.10.002, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=null, receivedDate=1743004800000, receivedDateStr=2025-03-27, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1774497648255, onlineDateStr=2026-03-26, pubDate=1760889600000, pubDateStr=2025-10-20, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1774497648255, onlineIssueDateStr=2026-03-26, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1774497648255, creator=13701087609, updateTime=1774497648255, updator=13701087609, issue=Issue{id=1243896547704947617, tenantId=1146029695717560320, journalId=1240685776644648972, year='2025', volume='29', issue='10', pageStart='1499', pageEnd='1662', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=null, createTime=1774497647453, creator=13701087609, updateTime=1774501638614, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1243913287902020448, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1243896547704947617, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1243913287902020449, tenantId=1146029695717560320, journalId=1240685776644648972, issueId=1243896547704947617, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=1514, endPage=1526, ext={EN=ArticleExt(id=1243896551546925768, articleId=1243896551068775110, tenantId=1146029695717560320, journalId=1240685776644648972, language=EN, title=Machine learning prediction method for wave loads on ship hull with missing experimental data, columnId=1241023037940748650, journalTitle=Journal of Ship Mechanics, columnName=Hydrodynamics, runingTitle=null, highlight=null, articleAbstract=

The precise determination of low-frequency wave loads on ship hulls is an indispensable cornerstone and core challenge in hull structural design. Low-cost ship model testing is often employed in engineering to forecast wave loads. However, test data frequently suffer from deficiencies or abnormalities due to various reasons. Consequently, predicting wave loads from data with defects or anomalies, remains a major engineering challenge. This paper presented an efficient method for accurately determining the wave design loads on ship hulls, specifically tailored to handle deficient or abnormal test data. By integrating 5,400 sets of wave load data calculated using two-dimensional strip theory, a machine learning transfer network was constructed. To address deficient data, we innovatively introduced a fine-tuning network layer, and designed a novel loss function that ignores zero terms, thereby enhancing the network's adaptability. This method achieved rapid wave load forecasting by transferring simulation results to ship model tests, with an accuracy better than 90%. This technique enhances design efficiency, reduces labor costs, and maximizes data utilization, providing a reliable and efficient solution for wave load prediction in hull structural design.

, correspAuthors=null, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=Zi-yuan WANG, Shuai CHEN, Cai-xia JIANG), CN=ArticleExt(id=1243896564293415922, articleId=1243896551068775110, tenantId=1146029695717560320, journalId=1240685776644648972, language=CN, title=针对缺损试验数据的船体波浪载荷机器学习预报方法, columnId=1241023038087549292, journalTitle=船舶力学, columnName=流体力学, runingTitle=null, highlight=null, articleAbstract=

船体低频波浪载荷的精确确定,构成了船体结构设计不可或缺的基石与核心挑战。工程中往往会采用船模试验进行波浪载荷预报。可试验数据往往会因各种原因产生缺损或异常,如何通过这些存在缺损或异常的试验数据对波浪载荷进行预报,是目前工程上的重要难题。本文提出了一种波浪载荷预测方法,针对缺损或异常的试验数据可高效且准确地预测水面船体波浪设计载荷。文中结合二维切片理论计算的5400组波浪载荷数据,构建了机器学习迁移网络,创新性地加入微调网络层处理缺损数据,同时设计新损失函数忽略零项,从而提升网络适应性。该方法通过迁移仿真结果至船模试验,实现了波浪载荷的快速预报,预报精度达到了90%以上。此技术显著提升了设计效率,降低了人力成本,并充分利用了现有数据资源,为船体结构设计提供了可靠且高效的波浪载荷预测方案。

, correspAuthors=null, authorNote=null, correspAuthorsNote=
通讯作者,E-mail:
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=wpzI1Dyi8sRFY83ZhKrWiA==, magXml=Ox+CK8cMuKGrN1DOFeiLLw==, pdfUrl=null, pdf=xHtxCTf4NRegnmPfEAGx/w==, pdfFileSize=2129432, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=uEOg3VnPHfKmiGX84uS2xw==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=8UlOd/T+6H8THkwZV0XHag==, mapNumber=null, authorCompany=null, fund=null, authors=

王子渊(1991-),男,硕士,高级工程师,通讯作者,E-mail:

, authorsList=王子渊, 陈帅, 蒋彩霞)}, authors=[Author(id=1243896564847063056, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=wangzy@cssrc.com.cn, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1243896564956114968, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, authorId=1243896564847063056, language=EN, stringName=Zi-yuan WANG, firstName=Zi-yuan, middleName=null, lastName=WANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1.China Ship Scientific Research Center, Wuxi 214082, China
2.Taihu Laboratory of Deepsea Technological Science, Wuxi 214082, China
3.Northwestern Polytechnical University, Xi’an 710072, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1243896565065166878, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, authorId=1243896564847063056, language=CN, stringName=王子渊, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, 2, 3, address=1.中国船舶科学研究中心,江苏 无锡 214082
2.深海技术科学太湖实验室,江苏 无锡 214082
3.西北工业大学,西安 710072, bio={"content":"

王子渊(1991-),男,硕士,高级工程师,通讯作者,E-mail:

"}, bioImg=null, bioContent=

王子渊(1991-),男,硕士,高级工程师,通讯作者,E-mail:

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1243896564532491260, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=1., ext=[AuthorCompanyExt(id=1243896564540879869, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1243896564545074174, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)]), AuthorCompany(id=1243896564658319363, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=2., ext=[AuthorCompanyExt(id=1243896564666707972, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564658319363, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Taihu Laboratory of Deepsea Technological Science, Wuxi 214082, China), AuthorCompanyExt(id=1243896564675096581, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564658319363, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.深海技术科学太湖实验室,江苏 无锡 214082)]), AuthorCompany(id=1243896564750594059, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=3., ext=[AuthorCompanyExt(id=1243896564758982667, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564750594059, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.Northwestern Polytechnical University, Xi’an 710072), AuthorCompanyExt(id=1243896564763176973, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564750594059, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.西北工业大学,西安 710072)])]), Author(id=1243896565149052961, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1243896565249716261, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, authorId=1243896565149052961, language=EN, stringName=Shuai CHEN, firstName=Shuai, middleName=null, lastName=CHEN, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1243896565337796651, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, authorId=1243896565149052961, language=CN, stringName=陈帅, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1243896564532491260, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=1., ext=[AuthorCompanyExt(id=1243896564540879869, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1243896564545074174, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])]), Author(id=1243896565425877037, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1243896565518151730, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, authorId=1243896565425877037, language=EN, stringName=Cai-xia JIANG, firstName=Cai-xia, middleName=null, lastName=JIANG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.China Ship Scientific Research Center, Wuxi 214082, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1243896565602037816, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, authorId=1243896565425877037, language=CN, stringName=蒋彩霞, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=1, address=1.中国船舶科学研究中心,江苏 无锡 214082, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1243896564532491260, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=1., ext=[AuthorCompanyExt(id=1243896564540879869, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1243896564545074174, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)])])], keywords=[Keyword(id=1243896565706895420, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, orderNo=1, keyword=transfer learning), Keyword(id=1243896565782392895, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, orderNo=2, keyword=fine tune), Keyword(id=1243896565891444803, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, orderNo=3, keyword=wave load prediction), Keyword(id=1243896565983719495, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, orderNo=4, keyword=missing experimental data), Keyword(id=1243896566055022666, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, orderNo=1, keyword=迁移学习), Keyword(id=1243896566151491662, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, orderNo=2, keyword=微调), Keyword(id=1243896566226989140, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, orderNo=3, keyword=波浪载荷预报), Keyword(id=1243896566315069527, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, orderNo=4, keyword=缺损试验数据)], refs=[Reference(id=1243896570807169256, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2021, volume=2, issue=null, pageStart=138, pageEnd=145, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=焦甲龙, 陈超核, 任慧龙, journalName=船舶力学, refType=null, unstructuredReference=焦甲龙,陈超核,任慧龙. 真实海况下船舶水弹性响应及砰击载荷的大尺度模型试验研究[J]. 船舶力学2021, 2:138-145., articleTitle=真实海况下船舶水弹性响应及砰击载荷的大尺度模型试验研究, refAbstract=null), Reference(id=1243896570895249644, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2021, volume=2, issue=null, pageStart=138, pageEnd=145, url=null, language=null, rfNumber=[1], rfOrder=1, authorNames=Jiao J L, Chen C H, Ren H L, journalName=Journal of Ship Mechanics, refType=null, unstructuredReference=Jiao J L, Chen C H, Ren H L. Investigation of hydroelastic response and slamming loads ofships in real waves by large-scale model measurement[J]. Journal of Ship Mechanics, 2021, 2:138-145. (in Chinese), articleTitle=Investigation of hydroelastic response and slamming loads ofships in real waves by large-scale model measurement, refAbstract=null), Reference(id=1243896570979135726, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2023, volume=10, issue=null, pageStart=1432, pageEnd=1444, url=null, language=null, rfNumber=[2], rfOrder=2, authorNames=陈帅, 蒋彩霞, 王子渊, journalName=船舶力学, refType=null, unstructuredReference=陈帅,蒋彩霞,王子渊,. 基于迁移学习的多源波浪载荷融合方法研究[J]. 船舶力学2023, 10:1432-1444., articleTitle=基于迁移学习的多源波浪载荷融合方法研究, refAbstract=null), Reference(id=1243896571071410416, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2023, volume=10, issue=null, pageStart=1432, pageEnd=1444, url=null, language=null, rfNumber=[2], rfOrder=3, authorNames=Chen S, Jiang C X, Wang Z Y, journalName=Journal of Ship Mechanics, refType=null, unstructuredReference=Chen S, Jiang C X, Wang Z Y, et al. Multi-source wave load fusion methodbased on transfer learning[J]. Journal of Ship Mechanics, 2023, 10:1432-1444. (in Chinese), articleTitle=Multi-source wave load fusion methodbased on transfer learning, refAbstract=null), Reference(id=1243896571155296498, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=4, authorNames=刘一夫, journalName=null, refType=null, unstructuredReference=刘一夫. 超大型集装箱船波浪载荷大尺度模型试验研究[D]. 哈尔滨:哈尔滨工程大学,2019., articleTitle=超大型集装箱船波浪载荷大尺度模型试验研究, refAbstract=null), Reference(id=1243896571222405363, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2019, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=5, authorNames=Liu Y F, journalName=null, refType=null, unstructuredReference=Liu Y F. Large scale model experimental study on wave load ofultra-large container ships[D]. Harbin: Harbin Engineering University, 2019. (in Chinese), articleTitle=Large scale model experimental study on wave load ofultra-large container ships, refAbstract=null), Reference(id=1243896571297902836, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2017, volume=28, issue=4, pageStart=39, pageEnd=46, url=null, language=null, rfNumber=[4], rfOrder=6, authorNames=肖武, 王海洋, journalName=船舶, refType=null, unstructuredReference=肖武,王海洋. 某大尺度新船型波浪载荷理论预报[J]. 船舶2017, 28(4):39-46., articleTitle=某大尺度新船型波浪载荷理论预报, refAbstract=null), Reference(id=1243896571390177524, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2017, volume=28, issue=4, pageStart=39, pageEnd=46, url=null, language=null, rfNumber=[4], rfOrder=7, authorNames=Xiao W, Wang H Y, journalName=Ship & Boat, refType=null, unstructuredReference=Xiao W, Wang H Y. Wave loads prediction of large scale new type ship[J]. Ship & Boat, 2017, 28(4): 39-46. (in Chinese), articleTitle=Wave loads prediction of large scale new type ship, refAbstract=null), Reference(id=1243896571486646519, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2017, volume=39, issue=15, pageStart=17, pageEnd=21, url=null, language=null, rfNumber=[5], rfOrder=8, authorNames=孙雪荣, 汪战军, 彭亚康, journalName=舰船科学技术, refType=null, unstructuredReference=孙雪荣,汪战军,彭亚康. 小方型系数船舶的船体波浪载荷综合研究[J]. 舰船科学技术2017, 39(15):17-21., articleTitle=小方型系数船舶的船体波浪载荷综合研究, refAbstract=null), Reference(id=1243896571578921206, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2017, volume=39, issue=15, pageStart=17, pageEnd=21, url=null, language=null, rfNumber=[5], rfOrder=9, authorNames=Sun X R, Wang Z J, Peng Y K, journalName=Ship Science and Technology, refType=null, unstructuredReference=Sun X R, Wang Z J, Peng Y K. The study of wave loads about small block coefficient ships[J]. Ship Science and Technology, 2017, 39(15): 17-21. (in Chinese), articleTitle=The study of wave loads about small block coefficient ships, refAbstract=null), Reference(id=1243896571692167417, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2015, volume=43, issue=8, pageStart=79, pageEnd=82, url=null, language=null, rfNumber=[6], rfOrder=10, authorNames=任慧龙, 陈亮亮, journalName=华中科技大学学报(自然科学版), refType=null, unstructuredReference=任慧龙,陈亮亮. 大型水面舰船波浪设计载荷的模型试验研究[J]. 华中科技大学学报(自然科学版)2015, 43(8):79-82., articleTitle=大型水面舰船波浪设计载荷的模型试验研究, refAbstract=null), Reference(id=1243896571847356668, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2015, volume=43, issue=8, pageStart=79, pageEnd=82, url=null, language=null, rfNumber=[6], rfOrder=11, authorNames=Ren H L, Chen L L, journalName=J. Huazhong Univ. of Sci. & Tech. (Natural Science Edition), refType=null, unstructuredReference=Ren H L, Chen L L. Model test investigation of design wave loads of large surface warship[J]. J. Huazhong Univ. of Sci. & Tech. (Natural Science Edition), 2015, 43(8): 79-82. (in Chinese), articleTitle=Model test investigation of design wave loads of large surface warship, refAbstract=null), Reference(id=1243896571939631358, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2008, volume=9, issue=null, pageStart=2458, pageEnd=2461, url=null, language=null, rfNumber=[7], rfOrder=12, authorNames=徐言民, 马国勤, journalName=系统仿真学报, refType=null, unstructuredReference=徐言民,马国勤. 大幅波浪中船舶非线性运动与波浪载荷仿真[J]. 系统仿真学报2008, 9:2458-2461., articleTitle=大幅波浪中船舶非线性运动与波浪载荷仿真, refAbstract=null), Reference(id=1243896572031906047, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2008, volume=9, issue=null, pageStart=2458, pageEnd=2461, url=null, language=null, rfNumber=[7], rfOrder=13, authorNames=Xu Y M, Ma G Q, journalName=Journal of System Simulation, refType=null, unstructuredReference=Xu Y M, Ma G Q. Simulation on nonlinear ship motions and wave loads with large amplitude waves[J]. Journal of System Simulation, 2008, 9:2458-2461. (in Chinese), articleTitle=Simulation on nonlinear ship motions and wave loads with large amplitude waves, refAbstract=null), Reference(id=1243896572107403522, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2008, volume=30, issue=6, pageStart=20, pageEnd=28, url=null, language=null, rfNumber=[8], rfOrder=14, authorNames=汪雪良, 顾学康, 祁恩荣, journalName=舰船科学技术, refType=null, unstructuredReference=汪雪良,顾学康,祁恩荣,. 船舶波浪载荷预报方法和模型试验研究综述[J]. 舰船科学技术2008, 30(6):20-28., articleTitle=船舶波浪载荷预报方法和模型试验研究综述, refAbstract=null), Reference(id=1243896572178706693, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2008, volume=30, issue=6, pageStart=20, pageEnd=28, url=null, language=null, rfNumber=[8], rfOrder=15, authorNames=Wang X L, Gu X K, Qi E R, journalName=Ship Science and Technology, refType=null, unstructuredReference=Wang X L, Gu X K, Qi E R, et al. A state-of-the-art review on study of prediction methods and model test of wave-induced loads of vessels[J]. Ship Science and Technology, 2008, 30(6): 20-28. (in Chinese), articleTitle=A state-of-the-art review on study of prediction methods and model test of wave-induced loads of vessels, refAbstract=null), Reference(id=1243896572258398471, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=1984, volume=null, issue=2, pageStart=69, pageEnd=82, url=null, language=null, rfNumber=[9], rfOrder=16, authorNames=徐能兹, 沈进威, 陈瑞章, journalName=中国造船, refType=null, unstructuredReference=徐能兹,沈进威,陈瑞章,. S-175集装箱船舶波浪载荷的船模试验与线性切片理论计算的比较[J]. 中国造船1984(2):69-82., articleTitle=S-175集装箱船舶波浪载荷的船模试验与线性切片理论计算的比较, refAbstract=null), Reference(id=1243896572342284554, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=1984, volume=null, issue=2, pageStart=69, pageEnd=82, url=null, language=null, rfNumber=[9], rfOrder=17, authorNames=Xu N Z, Shen J W, Chen R Z, journalName=Shipbuilding of China, refType=null, unstructuredReference=Xu N Z, Shen J W, Chen R Z, et al. Model tests of s-175 containership on wave loads and comparisons between experiments and calculations of linear strip theory[J]. Shipbuilding of China, 1984(2): 69-82. (in Chinese), articleTitle=Model tests of s-175 containership on wave loads and comparisons between experiments and calculations of linear strip theory, refAbstract=null), Reference(id=1243896572451336459, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2023, volume=35, issue=null, pageStart=117, pageEnd=140, url=null, language=null, rfNumber=[10], rfOrder=18, authorNames=Jiang C X, Liu Y B, Wang Z Y, journalName=Journal of Hydrodynamics, refType=null, unstructuredReference=Jiang C X, Liu Y B, Wang Z Y, et al. A multi-fidelity prediction model for vertical bending moment and total longitudinal stress of a ship based on composite neural network[J]. Journal of Hydrodynamics, 2023, 35:117-140., articleTitle=A multi-fidelity prediction model for vertical bending moment and total longitudinal stress of a ship based on composite neural network, refAbstract=null), Reference(id=1243896572564582670, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2019, volume=14, issue=Sup1, pageStart=63, pageEnd=90, url=null, language=null, rfNumber=[11], rfOrder=19, authorNames=Paik J K, Lee D H, Kim S J, journalName=Ships and Offshore Structures, refType=null, unstructuredReference=Paik J K, Lee D H, Kim S J, et al. A new method for determining the design values of wave-induced hull girder loads acting on ships[J]. Ships and Offshore Structures, 2019, 14(Sup1):63-90., articleTitle=A new method for determining the design values of wave-induced hull girder loads acting on ships, refAbstract=null), Reference(id=1243896572686217489, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2021, volume=167, issue=null, pageStart=103, pageEnd=902, url=null, language=null, rfNumber=[12], rfOrder=20, authorNames=Pena B, Huang L, journalName=Coastal Engineering, refType=null, unstructuredReference=Pena B, Huang L. Wave-gan: A deep learning approach for the prediction of nonlinear regular wave loads and run-up on a fixed cylinder[J]. Coastal Engineering, 2021, 167:103-902., articleTitle=Wave-gan: A deep learning approach for the prediction of nonlinear regular wave loads and run-up on a fixed cylinder, refAbstract=null), Reference(id=1243896572778492180, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, doi=null, pmid=null, pmcid=null, year=2018, volume=61, issue=null, pageStart=202, pageEnd=219, url=null, language=null, rfNumber=[13], rfOrder=21, authorNames=Kumar N K, Savitha R, Al Mamun A, journalName=Marine Structures, refType=null, unstructuredReference=Kumar N K, Savitha R, Al Mamun A. Ocean wave characteristics prediction and its load estimation on marine structures: A transfer learning approach[J]. Marine Structures, 2018, 61:202-219., articleTitle=Ocean wave characteristics prediction and its load estimation on marine structures: A transfer learning approach, refAbstract=null)], funds=[Fund(id=1243896570689728740, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, awardId=WDZC70202030201, language=CN, fundingSource=国防科工局稳定支持基金资助项目(WDZC70202030201), fundOrder=null, country=null)], companyList=[AuthorCompany(id=1243896564532491260, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=1., ext=[AuthorCompanyExt(id=1243896564540879869, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.China Ship Scientific Research Center, Wuxi 214082, China), AuthorCompanyExt(id=1243896564545074174, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564532491260, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=1.中国船舶科学研究中心,江苏 无锡 214082)]), AuthorCompany(id=1243896564658319363, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=2., ext=[AuthorCompanyExt(id=1243896564666707972, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564658319363, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.Taihu Laboratory of Deepsea Technological Science, Wuxi 214082, China), AuthorCompanyExt(id=1243896564675096581, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564658319363, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=2.深海技术科学太湖实验室,江苏 无锡 214082)]), AuthorCompany(id=1243896564750594059, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, xref=3., ext=[AuthorCompanyExt(id=1243896564758982667, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564750594059, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.Northwestern Polytechnical University, Xi’an 710072), AuthorCompanyExt(id=1243896564763176973, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, companyId=1243896564750594059, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=3.西北工业大学,西安 710072)])], figs=[ArticleFig(id=1243896566440898651, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.1, caption=Flow chart of transfer learning strategy, figureFileSmall=PlrwEW58207fARdlNvWGqg==, figureFileBig=uEOg3VnPHfKmiGX84uS2xw==, tableContent=null), ArticleFig(id=1243896566512201823, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图1, caption=迁移学习流程图, figureFileSmall=PlrwEW58207fARdlNvWGqg==, figureFileBig=uEOg3VnPHfKmiGX84uS2xw==, tableContent=null), ArticleFig(id=1243896566730305638, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.2, caption=Deep neural network structure, figureFileSmall=OxtN/eNnjcn367+mjBF/Fw==, figureFileBig=l7iB3h5d6uvtVkzm/jdTDA==, tableContent=null), ArticleFig(id=1243896566830968941, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图2, caption=神经网络结构, figureFileSmall=OxtN/eNnjcn367+mjBF/Fw==, figureFileBig=l7iB3h5d6uvtVkzm/jdTDA==, tableContent=null), ArticleFig(id=1243896566906466415, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.3, caption=Training curves of vertical bending moment for different models, figureFileSmall=x/anIk1QFHL0nXAT5BVjXg==, figureFileBig=EB2oQJ0gLHKmVLuKqOYDZQ==, tableContent=null), ArticleFig(id=1243896567002935409, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图3, caption=各模型垂向弯矩训练曲线图, figureFileSmall=x/anIk1QFHL0nXAT5BVjXg==, figureFileBig=EB2oQJ0gLHKmVLuKqOYDZQ==, tableContent=null), ArticleFig(id=1243896567082627189, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.4, caption=Training loss functions of vertical bending moment for different models, figureFileSmall=8sEwSqfyMQ92rb+xbPZz0A==, figureFileBig=S5gI3AMdE+O4QbaZ3giFow==, tableContent=null), ArticleFig(id=1243896567174901880, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图4, caption=各模型垂向弯矩训练损失函数对比, figureFileSmall=8sEwSqfyMQ92rb+xbPZz0A==, figureFileBig=S5gI3AMdE+O4QbaZ3giFow==, tableContent=null), ArticleFig(id=1243896567279759485, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.5, caption=Probability distribution of relative errors in vertical bending moment predictions by various models, figureFileSmall=vZYXYtMvrD7SWrenCkHA7Q==, figureFileBig=rlflIm0V/GI3uAnfbyHGDg==, tableContent=null), ArticleFig(id=1243896567376228481, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图5, caption=各模型垂向弯矩预测结果相对误差概率分布图(纵坐标为样本个数), figureFileSmall=vZYXYtMvrD7SWrenCkHA7Q==, figureFileBig=rlflIm0V/GI3uAnfbyHGDg==, tableContent=null), ArticleFig(id=1243896567468503170, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.6, caption=Architecture of transform network, figureFileSmall=pQ1E+C/Iu0OaN9cTHeTVRg==, figureFileBig=s0SnyX99BjZXuE9B7mhUEg==, tableContent=null), ArticleFig(id=1243896567573360774, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图6, caption=迁移网络结构, figureFileSmall=pQ1E+C/Iu0OaN9cTHeTVRg==, figureFileBig=s0SnyX99BjZXuE9B7mhUEg==, tableContent=null), ArticleFig(id=1243896567690801291, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.7, caption=Diagram of tanh, figureFileSmall=y8SvRL8C8PdFk90oT8h6HQ==, figureFileBig=XDLbVOFMoRwfbFFghxSfHw==, tableContent=null), ArticleFig(id=1243896567804047501, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图7, caption=tanh函数图像, figureFileSmall=y8SvRL8C8PdFk90oT8h6HQ==, figureFileBig=XDLbVOFMoRwfbFFghxSfHw==, tableContent=null), ArticleFig(id=1243896567904710802, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.8, caption=Diagram of derivatives of tanh, figureFileSmall=Fz1/RuksGrFZwQjfMKGZMA==, figureFileBig=zXJKY705BK7PWIHZrRx0+g==, tableContent=null), ArticleFig(id=1243896568013762707, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图8, caption=tanh导数图像, figureFileSmall=Fz1/RuksGrFZwQjfMKGZMA==, figureFileBig=zXJKY705BK7PWIHZrRx0+g==, tableContent=null), ArticleFig(id=1243896568097648790, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.9, caption=Pictures of the tank experiment, figureFileSmall=KeU1cYNMInTiuEMRGvkT0Q==, figureFileBig=ZDyd4sNrIdoLXI4KEM/MeA==, tableContent=null), ArticleFig(id=1243896568181534875, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图9, caption=船模试验照片, figureFileSmall=KeU1cYNMInTiuEMRGvkT0Q==, figureFileBig=ZDyd4sNrIdoLXI4KEM/MeA==, tableContent=null), ArticleFig(id=1243896568319946906, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.10, caption=Convergence of transform network, figureFileSmall=LGCKNSXfoBUZeT94k3A20w==, figureFileBig=Vzv/FIAdPFCwsTKAiHCVpw==, tableContent=null), ArticleFig(id=1243896568387055772, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图10, caption=迁移网络收敛度, figureFileSmall=LGCKNSXfoBUZeT94k3A20w==, figureFileBig=Vzv/FIAdPFCwsTKAiHCVpw==, tableContent=null), ArticleFig(id=1243896568483524768, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.11, caption=Wave load forecasting results of target ship of Condition 1, figureFileSmall=uTwhF0mrbsek7icVQjuxgA==, figureFileBig=eNn63dBuf2D9NVI8SNA0HA==, tableContent=null), ArticleFig(id=1243896568621936801, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图11, caption=目标船波浪载荷预报结果工况一, figureFileSmall=uTwhF0mrbsek7icVQjuxgA==, figureFileBig=eNn63dBuf2D9NVI8SNA0HA==, tableContent=null), ArticleFig(id=1243896568722600101, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.12, caption=Wave load forecasting results of target ship of Condition 2, figureFileSmall=SQzaw1/ekK4FtPWGPDe8yA==, figureFileBig=W/EBVux5STqD5xJfl1qEfg==, tableContent=null), ArticleFig(id=1243896568819069096, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图12, caption=目标船波浪载荷预报结果工况二, figureFileSmall=SQzaw1/ekK4FtPWGPDe8yA==, figureFileBig=W/EBVux5STqD5xJfl1qEfg==, tableContent=null), ArticleFig(id=1243896568898760875, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Fig.13, caption=Wave load forecasting results of target ship of Condition 3, figureFileSmall=y05sAj0gtUSZUm0zdK0X5w==, figureFileBig=S+qJ2hh87UqdBmOPukt9Pg==, tableContent=null), ArticleFig(id=1243896568974258349, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=图13, caption=目标船波浪载荷预报结果工况三, figureFileSmall=y05sAj0gtUSZUm0zdK0X5w==, figureFileBig=S+qJ2hh87UqdBmOPukt9Pg==, tableContent=null), ArticleFig(id=1243896569049755825, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.1, caption=

Parameters range of the training dataset

, figureFileSmall=null, figureFileBig=null, tableContent=
训练数据参数特征周期航速有义波高航向角
参数范围/单位3.5~18.5/s0~16.09/(m·s−10~16.5/m0~180/°
), ArticleFig(id=1243896569129447603, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表1, caption=

训练数据参数范围

, figureFileSmall=null, figureFileBig=null, tableContent=
训练数据参数特征周期航速有义波高航向角
参数范围/单位3.5~18.5/s0~16.09/(m·s−10~16.5/m0~180/°
), ArticleFig(id=1243896569213333685, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.2, caption=

Network structure parameters

, figureFileSmall=null, figureFileBig=null, tableContent=
网络模型网络参数网络层数隐藏层神经元数网络模型类型
Square_51+4+150×50×50×50全连接神经网络
Pyramid_51+4+1128×64×32×16全连接神经网络
CAF_51+2+4+1Conv(50×50×2)×Conv(50×50×2)×128×64×32×16卷积混合神经网络
LFWLIP1+4+1192×96×48×24全连接神经网络
), ArticleFig(id=1243896569288831159, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表2, caption=

网络结构参数

, figureFileSmall=null, figureFileBig=null, tableContent=
网络模型网络参数网络层数隐藏层神经元数网络模型类型
Square_51+4+150×50×50×50全连接神经网络
Pyramid_51+4+1128×64×32×16全连接神经网络
CAF_51+2+4+1Conv(50×50×2)×Conv(50×50×2)×128×64×32×16卷积混合神经网络
LFWLIP1+4+1192×96×48×24全连接神经网络
), ArticleFig(id=1243896569368522938, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.3, caption=

Statistics of the test results of vertical bending moment for different models

, figureFileSmall=null, figureFileBig=null, tableContent=
模型μσ20%误差内样本个数10%误差内样本个数5%误差内样本个数
Square_50.00600.0594515146013498
Pyramid_50.01740.0569496444383405
CAF_50.00880.0577511846403477
LFWLIP0.00630.0442519949464255
), ArticleFig(id=1243896569431437502, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表3, caption=

各模型垂向弯矩测试结果相对误差统计表

, figureFileSmall=null, figureFileBig=null, tableContent=
模型μσ20%误差内样本个数10%误差内样本个数5%误差内样本个数
Square_50.00600.0594515146013498
Pyramid_50.01740.0569496444383405
CAF_50.00880.0577511846403477
LFWLIP0.00630.0442519949464255
), ArticleFig(id=1243896569515323585, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.4, caption=

Architecture parameters of transform layer network

, figureFileSmall=null, figureFileBig=null, tableContent=
网络参数参数取值网络参数参数取值
网络层数3批量大小8
输入层神经元数3训练/验证/测试数据0.6/0.3/0.1
输出层神经元数3优化算法Adam
隐藏层神经元数32激活函数Sigmond
学习率0.001最大训练次数100 000
), ArticleFig(id=1243896569607598276, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表4, caption=

迁移层网络结构参数

, figureFileSmall=null, figureFileBig=null, tableContent=
网络参数参数取值网络参数参数取值
网络层数3批量大小8
输入层神经元数3训练/验证/测试数据0.6/0.3/0.1
输出层神经元数3优化算法Adam
隐藏层神经元数32激活函数Sigmond
学习率0.001最大训练次数100 000
), ArticleFig(id=1243896569683095751, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.5, caption=

Parameters of target ship

, figureFileSmall=null, figureFileBig=null, tableContent=
船型参数/单位参数取值船型参数/单位参数取值
总长/m311设计吃水/m9.8
型深/m27设计排水量/t67 000
设计水线宽/m36.2方形系数0.613
), ArticleFig(id=1243896569779564745, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表5, caption=

目标船参数

, figureFileSmall=null, figureFileBig=null, tableContent=
船型参数/单位参数取值船型参数/单位参数取值
总长/m311设计吃水/m9.8
型深/m27设计排水量/t67 000
设计水线宽/m36.2方形系数0.613
), ArticleFig(id=1243896569880228045, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.6, caption=

Cases for comparison

, figureFileSmall=null, figureFileBig=null, tableContent=
工况航速/m·s−1特征周期/s有义波高/m航向角/°
工况一9.26212.714135
工况二5.14512.714180
工况三9.26212.2518.5120
), ArticleFig(id=1243896569980891341, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表6, caption=

工况表

, figureFileSmall=null, figureFileBig=null, tableContent=
工况航速/m·s−1特征周期/s有义波高/m航向角/°
工况一9.26212.714135
工况二5.14512.714180
工况三9.26212.2518.5120
), ArticleFig(id=1243896570068971728, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.7, caption=

Wave load prediction results of target ship of Condition 1

, figureFileSmall=null, figureFileBig=null, tableContent=
站位试验值2D-ST计算误差TL预报误差MANN预报误差
52269.42870.626.5%2138.95.7%2366.314.2%
106365.83310.248.0%6494.52.0%6875.728.0%
14.53661.11231.766.3%3778.93.2%3893.636.3%
), ArticleFig(id=1243896570178023635, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表7, caption=

目标船波浪载荷预报结果工况一

, figureFileSmall=null, figureFileBig=null, tableContent=
站位试验值2D-ST计算误差TL预报误差MANN预报误差
52269.42870.626.5%2138.95.7%2366.314.2%
106365.83310.248.0%6494.52.0%6875.728.0%
14.53661.11231.766.3%3778.93.2%3893.636.3%
), ArticleFig(id=1243896570253521111, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.8, caption=

Wave load forecasting results of target ship of Condition 2

, figureFileSmall=null, figureFileBig=null, tableContent=
站位试验值2D-ST计算误差TL预报误差MANN预报误差
5(缺失)2919.22604.42508.6
106472.64125.436.2%5865.29.3%6311.42.4%
14.53913.91460.662.6%3873.11.5%4263.48.9%
), ArticleFig(id=1243896570349990107, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表8, caption=

目标船波浪载荷预报结果工况二

, figureFileSmall=null, figureFileBig=null, tableContent=
站位试验值2D-ST计算误差TL预报误差MANN预报误差
5(缺失)2919.22604.42508.6
106472.64125.436.2%5865.29.3%6311.42.4%
14.53913.91460.662.6%3873.11.5%4263.48.9%
), ArticleFig(id=1243896570463236317, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=EN, label=Tab.9, caption=

Wave load forecasting results of target ship of Condition 3

, figureFileSmall=null, figureFileBig=null, tableContent=
站位试验值2D-ST计算误差TL预报误差MANN预报误差
52391.82883.120.5%2440.02.1%2259.65.53%
106482.63901.739.8%6350.02.0%6305.62.73%
14.53955.91542.061.0%4115.94.1%3721.35.93%
), ArticleFig(id=1243896570547122398, tenantId=1146029695717560320, journalId=1240685776644648972, articleId=1243896551068775110, language=CN, label=表9, caption=

目标船波浪载荷预报结果工况三

, figureFileSmall=null, figureFileBig=null, tableContent=
站位试验值2D-ST计算误差TL预报误差MANN预报误差
52391.82883.120.5%2440.02.1%2259.65.53%
106482.63901.739.8%6350.02.0%6305.62.73%
14.53955.91542.061.0%4115.94.1%3721.35.93%
)], attaches=null, journal=Journal(id=1240685663704625163, delFlag=0, nameCn=船舶力学, nameEn=Journal of Ship Mechanics, nameHistory1=null, nameHistory2=null, issn=1007-7294, eissn=null, cn=32-1468/U, coden=null, periodic=0, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=8VABIsgSw5I4HF/4goaIlA==, journalPrice=null, startedYear=null, abbrevIsoEn=Journal of Ship Mechanics, journalRemark=null, publicationField=null, createdTime=1773732113054, updatedTime=1773732336269, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=J, firstLetterEn=J, subjectCode=Engineering, subjectName=null, subjectCodeEn=Engineering, subjectNameEn=null, picCn=8VABIsgSw5I4HF/4goaIlA==, picEn=Ve/8A/FRYceS/z8o7GCJ5g==, jcr=null, cjcr=null, exts=[JournalExt(id=1240686600003318325, language=CN, name=船舶力学, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1773732336283, updatedTime=1773732336283, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=1, submissionEditorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=3, submissionReviewUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1240686600041067062, language=EN, name=Journal of Ship Mechanics, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1773732336292, updatedTime=1773732336292, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=1, submissionEditorUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=3, submissionReviewUrl=https://cblx.cbpt.cnki.net/EditorD3N/index.aspx?t=2, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1240685776644648972, websiteList=[Website(id=1240686760611607145, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1240685776644648972, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/cblx/CN, language=CN, createTime=1773732374576, createBy=18614031015, updateTime=1773732394922, updateBy=18614031015, name=船舶力学-中文, tplId=1146099689490845704, title=船舶力学, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1240687166741869216, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=articleTextType, value=kx, createTime=1773732471404, updateTime=1773732471404, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166725091997, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=banner, value=null, createTime=1773732471400, updateTime=1773732471400, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166758646435, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=grayFlag, value=0, createTime=1773732471408, updateTime=1773732471408, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166716703388, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=logo, value=https://castjournals.cast.org.cn/joweb/cblx/CN/file/pic?fileId=/ZVjg3ao2vf5wYxw5Kn1Uw==, createTime=1773732471398, updateTime=1773732471398, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166771229349, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=minRunFlag, value=0, createTime=1773732471411, updateTime=1773732471411, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166737674911, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/cblx/CN/file/pic, createTime=1773732471403, updateTime=1773732471403, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166767035044, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=silenceFlag, value=0, createTime=1773732471410, updateTime=1773732471410, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166729286302, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1773732471401, updateTime=1773732471401, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166750257825, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=themeColor, value=null, createTime=1773732471406, updateTime=1773732471406, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687166754452130, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760611607145, code=themeStyle, value=null, createTime=1773732471407, updateTime=1773732471407, creator=18614031015, updator=18614031015)]), Website(id=1240686760674521707, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1240685776644648972, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/cblx/EN, language=EN, createTime=1773732374590, createBy=18614031015, updateTime=1773732413338, updateBy=18614031015, name=船舶力学-英文, tplId=1146101810881728533, title=Journal of Ship Mechanics, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1240687194269078438, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=articleTextType, value=kx, createTime=1773732477967, updateTime=1773732477967, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194248106915, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=banner, value=null, createTime=1773732477962, updateTime=1773732477962, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194285855657, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=grayFlag, value=0, createTime=1773732477971, updateTime=1773732477971, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194239718306, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=logo, value=https://castjournals.cast.org.cn/joweb/cblx/EN/file/pic?fileId=/ZVjg3ao2vf5wYxw5Kn1Uw==, createTime=1773732477960, updateTime=1773732477960, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194302632875, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=minRunFlag, value=0, createTime=1773732477975, updateTime=1773732477975, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194260689829, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/cblx/EN/file/pic, createTime=1773732477965, updateTime=1773732477965, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194294244266, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=silenceFlag, value=0, createTime=1773732477973, updateTime=1773732477973, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194256495524, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1773732477964, updateTime=1773732477964, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194277467047, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=themeColor, value=null, createTime=1773732477969, updateTime=1773732477969, creator=18614031015, updator=18614031015), WebsiteProps(id=1240687194281661352, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1240686760674521707, code=themeStyle, value=null, createTime=1773732477970, updateTime=1773732477970, creator=18614031015, updator=18614031015)])], journalTitle=船舶力学, weixinUrl=null, journalUrl=https://cblx.cbpt.cnki.net/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Journal of Ship Mechanics, journalPhotoCn=8VABIsgSw5I4HF/4goaIlA==, journalPhotoEn=Ve/8A/FRYceS/z8o7GCJ5g==, journalFirstLetter=J, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/cblx/CN/10.3969/j.issn.1007-7294.2025.10.002, detailUrlEn=https://castjournals.cast.org.cn/joweb/cblx/EN/10.3969/j.issn.1007-7294.2025.10.002, pdfUrlCn=https://castjournals.cast.org.cn/joweb/cblx/CN/PDF/10.3969/j.issn.1007-7294.2025.10.002, pdfUrlEn=https://castjournals.cast.org.cn/joweb/cblx/EN/PDF/10.3969/j.issn.1007-7294.2025.10.002, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
针对缺损试验数据的船体波浪载荷机器学习预报方法
收藏切换
PDF下载
王子渊 1, 2, 3 , 陈帅 1 , 蒋彩霞 1
船舶力学 | 流体力学 2025,29(10): 1514-1526
收起
收藏切换
船舶力学 | 流体力学 2025, 29(10): 1514-1526
针对缺损试验数据的船体波浪载荷机器学习预报方法
全屏
王子渊1, 2, 3 , 陈帅1, 蒋彩霞1
作者信息
  • 1.中国船舶科学研究中心,江苏 无锡 214082
  • 2.深海技术科学太湖实验室,江苏 无锡 214082
  • 3.西北工业大学,西安 710072
  • 王子渊(1991-),男,硕士,高级工程师,通讯作者,E-mail:

通讯作者:

通讯作者,E-mail:
Machine learning prediction method for wave loads on ship hull with missing experimental data
Zi-yuan WANG1, 2, 3 , Shuai CHEN1, Cai-xia JIANG1
Affiliations
  • 1.China Ship Scientific Research Center, Wuxi 214082, China
  • 2.Taihu Laboratory of Deepsea Technological Science, Wuxi 214082, China
  • 3.Northwestern Polytechnical University, Xi’an 710072
出版时间: 2025-10-20 doi: 10.3969/j.issn.1007-7294.2025.10.002
文章导航
收藏切换

船体低频波浪载荷的精确确定,构成了船体结构设计不可或缺的基石与核心挑战。工程中往往会采用船模试验进行波浪载荷预报。可试验数据往往会因各种原因产生缺损或异常,如何通过这些存在缺损或异常的试验数据对波浪载荷进行预报,是目前工程上的重要难题。本文提出了一种波浪载荷预测方法,针对缺损或异常的试验数据可高效且准确地预测水面船体波浪设计载荷。文中结合二维切片理论计算的5400组波浪载荷数据,构建了机器学习迁移网络,创新性地加入微调网络层处理缺损数据,同时设计新损失函数忽略零项,从而提升网络适应性。该方法通过迁移仿真结果至船模试验,实现了波浪载荷的快速预报,预报精度达到了90%以上。此技术显著提升了设计效率,降低了人力成本,并充分利用了现有数据资源,为船体结构设计提供了可靠且高效的波浪载荷预测方案。

迁移学习  /  微调  /  波浪载荷预报  /  缺损试验数据

The precise determination of low-frequency wave loads on ship hulls is an indispensable cornerstone and core challenge in hull structural design. Low-cost ship model testing is often employed in engineering to forecast wave loads. However, test data frequently suffer from deficiencies or abnormalities due to various reasons. Consequently, predicting wave loads from data with defects or anomalies, remains a major engineering challenge. This paper presented an efficient method for accurately determining the wave design loads on ship hulls, specifically tailored to handle deficient or abnormal test data. By integrating 5,400 sets of wave load data calculated using two-dimensional strip theory, a machine learning transfer network was constructed. To address deficient data, we innovatively introduced a fine-tuning network layer, and designed a novel loss function that ignores zero terms, thereby enhancing the network's adaptability. This method achieved rapid wave load forecasting by transferring simulation results to ship model tests, with an accuracy better than 90%. This technique enhances design efficiency, reduces labor costs, and maximizes data utilization, providing a reliable and efficient solution for wave load prediction in hull structural design.

transfer learning  /  fine tune  /  wave load prediction  /  missing experimental data
王子渊, 陈帅, 蒋彩霞. 针对缺损试验数据的船体波浪载荷机器学习预报方法. 船舶力学, 2025 , 29 (10) : 1514 -1526 . DOI: 10.3969/j.issn.1007-7294.2025.10.002
Zi-yuan WANG, Shuai CHEN, Cai-xia JIANG. Machine learning prediction method for wave loads on ship hull with missing experimental data[J]. Journal of Ship Mechanics, 2025 , 29 (10) : 1514 -1526 . DOI: 10.3969/j.issn.1007-7294.2025.10.002
船体低频波浪载荷的确定是船体结构设计的前提与关键,如何获得不同结构类型船舶的外载荷是目前国内外的研究热点与难点。一般通过规范计算、数值计算、模型试验等方法获得结构外载荷,焦甲龙等[1]通过建立大尺度模型试验,研究了真实海况下船舶水弹性响应及砰击载荷。陈帅等[2]基于迁移学习,根据波浪载荷模型试验数据仓库建立的船体波浪载荷神经网络代理模型,提出了一种多源数据融合的波浪载荷预测方法,能简单、快捷、高效、相对准确地确定船体的波浪设计载荷。刘一夫[3]研究了超大型集装箱船波浪载荷,采用大尺度模型试验,分析运动与载荷响应,评估船舶安全性,为实船设计提供依据,并验证了大尺度模型试验技术的优越性。肖武等[4]运用专业软件对某大尺度新船型的波浪载荷进行了理论预报,并与船模试验结果进行了对比分析,通过数据推导,验证了非线性预报方法的有效性,为船舶设计提供了重要参考。孙雪荣等[5]结合理论与模型试验,综合研究了小方型系数船舶在复杂海况下的波浪载荷,通过数据推导揭示了该类船舶波浪载荷的非线性特性,并提出了规范修正建议。任慧龙等[6]针对大型水面舰船,通过模型试验推导了其波浪设计载荷,试验结果表明,现行规范对砰击弯矩的考虑不足,为波浪载荷规范修订提供了实验依据。徐言民等[7]基于切片理论扩展,考虑湿表面等非线性因素,开发了船舶非线性波浪载荷仿真平台,通过船模试验验证精度,并进行了大幅波浪中船舶非线性运动与载荷仿真。汪雪良等[8]调研了船舶波浪载荷预报的二维、三维方法及水弹性力学分析方法,分析了完全弹性及分段弹性模型试验技术,探讨了预报方法和模型试验的现状与发展趋势。徐能兹等[9]对比了S-175集装箱船波浪载荷的船模试验与线性切片理论计算结果,发现理论计算在波浪载荷纵向分布及中拱中垂分量上与试验值差异大,建议开展非线性理论与试验研究。从上述研究结果可以看出,只依赖于规范计算或数值计算预报船体波浪载荷,会存在较大误差,然而模型试验不仅成本高昂,且工况有限,因此,亟需引入最新的技术手段,对船体波浪载荷进行预报。机器学习作为近年来的热门方法,被广泛运用于船舶波浪载荷预报中。Jiang等[10]基于一种复合神经网络,预测了船体垂向波浪载荷。Paik等[11]提出了一种基于概率来确定船体主梁上波浪诱导载荷设计值的新方法,在实例船的试验中,该方法表现出了同船级社规则方法计算结果的一致性。Pena等[12]基于生成对抗网络原理提出了一种新的数据驱动计算技术Wave-GAN,该网络学习训练所用到的数据集是使用CFD模拟构建的,预测结果表明,Wave-GAN能够以较高的精度复刻CFD的结果,同时极大地降低了时间计算成本。Kumar等[13]利用迁移学习方法对12种不同海洋结构物的波浪诱导载荷进行了预测。这些方法都对提高设计效率、节省人力成本、利用现有数据资源等方面具有重要的意义。
在实际工程中,试验数据的采集和处理是一项至关重要的任务。然而,由于多种内外因素的影响,试验数据往往会出现缺损或测量异常的情况。这些因素可能包括但不限于设备故障、环境干扰、操作失误等。这种数据的不完整性或者异常性,对于后续的数据分析和波浪载荷预报带来了极大的挑战。如何通过这些夹杂着异常数据的试验结果,准确地对波浪载荷进行预报,是目前工程实践中亟待解决的一个重要技术问题。这不仅需要高效的数据清洗和预处理方法,还需要借助先进的统计和机器学习技术,以便从复杂的噪声数据中提取出有价值的信息,从而实现对波浪载荷的准确预报。
本文整理了邮船波浪载荷物理试验数据,并利用二维切片理论计算了5400组不规则波波浪载荷算例,并将这些算例用于构建机器学习DNN网络,形成迁移网络中的源域,从而将仿真计算的波浪载荷预报的知识迁移至船模试验中,实现对船模试验下的波浪载荷快速预报。本文将以邮船为例,介绍一种创新的迁移学习方法,该方法通过引入一个独立的微调网络层,构建一种全新的损失函数,引入激活函数,使得缺损数据在反向传递过程中不参与计算,从而有效地解决因缺损试验数据导致的机器学习模型训练中的关键问题,即输入秩不一致无法进行计算的情况。
船舶运动和载荷预报的传统方法是基于细长体假设的切片法(2D-Strip Theory,2D-ST),采用平面流假设,把任意的船体横剖面看成无限长柱状体的一部分,将船体周围的三维流动转化为剖面内的二维流动问题来求解船体周围的流场,由此得出剖面作升沉、横荡和横摇运动的二维流体动力,沿船长积分得出三维流体动力,从而得出船体的波浪载荷。该方法对常规细长船舶的运动响应和剖面波浪载荷能给出较好的预报结果。对于常规单体船型,如散货船、邮船等,在中低航速下,细长体假设能够较好地反映船舶的流体动力学特性。
船体“切片”,即单位长度船体上的刚性惯性力载荷f1可表示为:
式中,µjGXh分别表示单位长度船体的质量、对质心轴CX的转动惯量、质心C与船体重心G的垂向距离,yG为纵向位移,ZG为垂向位移,θx为绕x轴转角,θy为绕y轴转角,θz为绕z轴转角。
全船的质量M和重心轴的转动惯量JGXJGX可表示为:
记船体横剖面内的力和矩的6个分量依次为轴力N、垂向剪力SFZ、弯矩BMY、水平剪力SFY、弯矩BMZ和扭矩TMG,通常轴力N很小,可以忽略不计。上述力分量分别表示为:
式中,Ia为重心G至船尾的纵向距离。不难验证剪力、弯矩和扭矩的封闭条件。
波浪载荷预报的目的是以规则波中的波浪载荷响应为基础,通过理论计算,确定船舶在给定的时间范围运行于实际海况中的波浪载荷变化特性。
波浪载荷短期预报的时间范围为半小时到数小时,在此期间,船舶的装载状态、航速、航向以及海况都可以认为是固定不变的(在统计意义上)。短期海浪可视为均值为零的平稳正态随机过程,当船体对波浪的响应是线性系统时,由随机过程理论可知,船舶的响应与波浪同属于平稳随机过程。波浪和响应之间的关系为:
式中,SζωH1/3TZθ)是海浪谱密度;SWωH1/3TZVβ+θ)是波浪载荷的谱密度;HωVTZβ+θ)是系统传递函数的模,其值为单位规则波下的载荷响应幅值;ω是波浪圆频率;V是航速;θ是组合波与主浪向的夹角;H1/3是有义波高;TZ是波浪的特征周期;β是航向角。
但在计算中,实际频率应是遭遇频率,所以输入的波浪谱SζωH1/3TZθ)应变换为SζωeH1/3TZθ),根据对于频率微元间的能量不变关系,两者间有:
深水中的遭遇频率ωe与波浪频率ω之间的关系是:
由此得到遭遇频率下的波浪谱为:
由概率论可知,对于一个均值为零的平稳高斯过程,在响应谱是窄谱的假设下,其幅值X服从Rayleigh分布,对应的概率密度为:
式中,σ2=DM=m0。这个参数可由载荷响应谱SWωH1/3TZVβ+θ)得到。
由谱密度函数的性质,可知方差为:
由此可以得到波浪载荷的各种特征值,例如对单幅值有:
均值;有义值;1/10最大平均值
当前的人工智能技术大多需要有大量高质量的数据支撑,使用实验室生成的模拟数据可以一定程度上解决这一难题,满足基本的训练需求。然而,船模试验的工况数目并不能满足机器学习的需求,实际拥有的数据不足,且工况以及剖面位置都十分有限,导致整船剖面的载荷预测结果不够准确。这一问题的出现提出了新的要求,即在充分利用船模试验的基础上,还要获得高精度的整船载荷预报结果。本文采用迁移学习(Transfer Learning,TL)技术作为一种解决方法,在迁移网络中加入独立的微调网络层,对于缺失的试验数据置零,并修改传递函数忽略零项,实现网络微调。整个预报流程如图1所示。
迁移学习过程中会把已经训练好的模型参数迁移到新的模型中来帮助其训练。考虑到大部分数据或任务是存在相关性的,所以通过迁移学习,可以将已经学到的模型参数通过某种方式分享给新模型从而加快并优化模型的学习效率。
迁移学习的基本动机是试图将从一个问题中获取的知识应用到另外一个不同但是相关的问题中去。本文将第1节快速生成的二维切片理论不规则波波浪载荷算例进行机器学习网络构建,建立预训练集,形成迁移网络中的源域,然后将船模试验数据进行迁移学习,实现对船模试验下的波浪载荷快速预报。
在开发迁移学习模型之前,选择高质量的数据集对于确保预测结果的准确性和稳定性至关重要。鉴于全面测试数据的有限性,本研究利用的是一个包含5400条针对特定船型的模拟数据条目的数据集。本文所采用的海况数据,是基于北大西洋波浪散布图,数据涵盖了该图中出现的所有波高和周期。所采用的不规则波谱是JONSWAP谱,因为JONSWAP谱更适合表达海浪非线性特征。有关数据集中所用海况的参数范围,如表1所示。
基于第1节所建立的二维切片理论不规则波波浪载荷预报方法,本研究针对多型邮轮开展不规则波条件下的短期载荷预报。采用拉丁超立方抽样(Latin Hypercube Sampling,LHS)技术对输入参数空间进行实验设计,选取的输入参数包括波浪参数(波高、周期、航向角)与船体参数(航速、站位),共同构成多维设计空间。
拉丁超立方抽样作为一种分层抽样技术,其核心优势在于通过等概率区间划分实现设计空间的高效覆盖。该方法具有优良的空间填充特性与概率映射能力,可确保样本点在多维参数空间内呈现均匀分布状态。
拉丁超立方抽样的实施流程包含以下三个关键步骤:
(1)维度离散化处理:对每个设计参数维度进行m个等概率区间划分,区间长度根据该维度参数的概率密度函数进行自适应调整(当参数服从均匀分布时,各区间长度相等);
(2)样本矩阵构建:在各维度的每个子区间内随机生成一个样本点,形成m×n维样本矩阵(n为设计参数维度数),理论上可产生mn个候选样本点;
(3)样本空间优化:采用无放回抽样策略从候选集中抽取m个样本向量,确保每个维度上的每个子区间仅被采样一次,从而实现样本在多维空间内的最优分布。
本研究通过实施拉丁超立方抽样技术构建了包含5400组参数组合的样本数据库,结合二维切片理论开发的数值计算程序进行波浪载荷求解,最终输出垂向弯矩、横向弯矩及扭矩等关键载荷参数。
在五个特征(航速、周期、波高、航向角和站位)作为输入层的情况下,经过初步实验的筛选后,确定了三个表现相对较好的模型,分别是Square_5、Pyramid_5和CAF_5,再加上了中国船舶科学研究中心陈帅等[12]基于AutoML进行超参数优化形成的预报网络模型LFWLIP,现主要对这4个模型进行对比。4个模型均采用五个特征作为输入层、弯矩(或扭矩)作为输出层的结构,其主要不同在于中间隐藏层,对应的神经网络结构如图2所示。
Square_5的隐藏层是由四层全连接层组成的矩形网络,每层包含50个节点;Pyramid_5的隐藏层是由四层全连接层组成的倒金字塔形网络,每层节点个数分别为128、64、32、16;CAF_5的隐藏层由卷积层和全连接层混合而成,前两层为卷积层,每层50个卷积核,后面紧接着四层倒金字塔形的全连接层,每层节点同样分别为128、64、32、16个;LFWLIP的隐藏层与Pyramid_5类似是一个倒金字塔形网络,由四层全连接层组成的倒金字塔形网络,但每层节点数为192、96、48、24。4个模型的具体参数如表2所示。
4个模型的训练曲线如图3所示,其中横坐标为循环次数,纵坐标为损失函数误差。
图3中可以看出,各模型的loss曲线在100次循环以后均变得很平缓,说明模型已经无法通过增加训练次数再进一步优化。且图3(d)中验证集的曲线波动最小,表明其训练的模型在验证集上的预测结果最为稳定;相较之下,图3(c)代表的模型的稳定性不够好,预测结果的起伏相对较大。图4中分别比较了100次循环后,各个网络模型的训练误差曲线。
表3是各模型在垂向弯矩测试集上的预测结果统计,表格后面三列分别代表预测结果相对误差在20%以内、10%以内和5%以内的数据个数。从表中容易看出,Square_5、Pyramid_5和CAF_5三个模型预测结果相对误差的标准差以及在各区间的分布个数并没有很大的差别,同时观察图5(a)~(c),三个模型相对误差对应的概率分布图也很相似,说明这三个模型达到的预测效果基本一样。但是,此三个模型相对误差在5%以内的预测点个数均不到4000个,预测的精度还不够,而基于贝叶斯超参数优化(Bayesian Optimization,BO)后的网络模型LFWLIP则较之前的三个模型在预测精度上有了很大的提升,相对误差在5%以内的点个数增加了15%左右。图5(d)中的相对误差概率分布也明显优于其他三者,体现出了较高的精度。因此,本文采用了LFWLIP作为迁移预训练网络。针对网络结构参数对模型性能的影响,训练集、验证集和测试集数据划分为0.8/0.1/0.1,最大训练次数为20 000。
本文采用冻结迁移预训练网络参数、增加迁移层的方式进行迁移学习,原因是微调迁移层参数会降低模型可解释性,而保留迁移层网络架构与参数有助于对知识的保留与传递。我们将迁移层与预训练网络分开,之后可以对预训练层增加新的船型仿真训练数据,也可以对迁移层增加新的船模试验数据,两者互相不干扰,实现模型随着数据的积累从而形成更高精度的自进化。另一方面,模型试验数据往往并不像迁移网络输出那么完整,有时候对应输出项会减少,从而使输出矩阵异秩,整个网络无法直接进行逆向传播,因此面对这种数据,也需要单独建立迁移层,采用微调策略建立输出层,实现逆向传播。整个迁移学习网络构架如图6所示。
迁移层为全连接神经网络,网络架构参数如表4所示。
在现实世界的工程项目中,试验数据的搜集与处理扮演着至关重要的角色。然而,试验数据常因设备故障、操作不当、环境干扰等因素出现缺失或异常。这种情况对波浪载荷的精确预测构成了显著挑战。同时,数据的丢失导致迁移层网络与预训练网络在输入与输出上出现不匹配,使得在微调阶段无法进行有效的反向传播,进而导致预测模型的失效。因此,针对不完整的试验数据模型,本文将提出一种新的损失函数来应对这一问题。
为了保证预训练网络的输出和迁移层的输入一致,在迁移之前,先将不完整的试验输出数据即缺损或不正常的试验数据置为零,然后在损失函数中,通过构建一种新的均方根误差形式,在均方差公式中引入激活函数tanh函数。
tanh函数定义为:
其函数图象如图7所示。
该函数导数为:
其导数图像如图8所示。
通过其函数图像及导数图像可以看出,此种类sigmoid激活函数具有连续性,且在接近0时,函数快速归0;而在远离0时,函数结果快速归1,因此可以利用此特性将之前置零的异常或损失数据项快速从正常试验数据中挑出,具体操作如下。
为了只考虑非零项数据,使误差数据反向传播,这里参照均方根误差(Mean Squared Error,MSE)构建一个新的损失函数LSMSE为:
式中,o*为目标输出(即试验数据)中的非零项,这里对试验数据进行如下处理:
这里为了避免过小的试验数据也被作为置零项,α为扩大系数,因此扩大α倍以后代入激活函数,确保其值为1。(这里扩大系数基于试验数据数量级,就波浪载荷数据而言扩大1000倍可满足计算要求)
式(14)中p*为目标输出(即试验数据)中非零项所对应的预测值。该预测值在微调层第一次循环时为迁移层输出,该输出是没有置零项的,因此这里需要将其对应的缺失试验数据项置零,方法同上。
这里利用tanh函数的置零功能,将缺失试验数据对应的预测值项置零。
式(14)中n*为非零项的个数,其值为:
通过上述方法建立损失函数为:
通过上式就可以将存在缺失数据的试验数据流进行逆向传递,并且可以实现自我检测异常数据,从而无需在整个机器学习网络程序中对每条数据插入if语句进行检测,大大提高了机器学习网络的运算速度。
目标船为一艘邮船(船型参数见表5),中国船舶科学研究中心针对目标船的波浪载荷进行了船模水池试验,测量了各个航态下的垂向波浪弯矩、水平波浪弯矩和扭矩等外载荷。
试验模型如图9所示,其缩尺比为1∶40,采用玻璃钢材料制作而成。试验模型沿着主船体方向分为3段,各段经称重处理,其横剖面梁的刚度与主船体中横剖面垂向抗弯刚度一致,用测量梁相互连接成一个整体,并用胶皮连接防止进水,各个分段端部设有横舱壁。船模在波浪中自航运动时,在分段间隙处通过测量梁上布置的应变传感器测量船体梁的波浪弯矩动响应。
由于水池试验过程中存在某些工况数据丢失或异常情况,并且只能分成3个段位,而正常船舶载荷计算通常需要20个段位数据。因此,将空白段位数据及传感器异常数据置零,进入网络进行训练计算。
对于迁移预训练网络结构,由于训练集数据量较大,如第2章所示,训练数据得到了很好的收敛性,在200步训练以后,模型预测LOSS值降至0.001以下,说明整个网络构建合适(如图10所示)。
对于迁移层,由于训练数据较少,且与模型试验数据和二维线性仿真结果存在较大的差别,整个训练模型收敛性低于预训练模型。
表6列出了基于波高、航向角与航速选取的3个典型试验工况。浙江大学刘玉波等[6]基于多保真数据回归网络(Multi-fidelity artificial neural network,MANN)也对该船进行了波浪载荷预报,本文同时得出该方法计算结果以进行对比。
图11~13表7~9,同时对比分析了二维切片仿真计算、本文迁移学习的波浪载荷预报以及基于MANN的预报结果。
通过对比可以看出,本文所使用的迁移学习方法在缺损试验数据的情况下,精度与浙大的MANN方法相当,且远高于仿真计算。但是浙大的MANN方法处理缺损试验数据的方式是对每个站位数据进行独立标签,将其作为一个独立工况进行计算,因此,计算量是本文方法的二十倍,计算时间也大幅增加。并且从工况二可以看出,MANN方法对于存在数据点位置的地方精度较高,但整个趋势与过往经验数据以及理论相比,其线型较窄,说明峰值对其预报影响较大,其模型泛化性较本文所使用的方法要差,本文所采用的方法预报出来的载荷趋势更符合真实情况。
本文提出的波浪载荷预测方法相较于传统仿真技术,能够实现更高精度地波浪载荷预测。该方法的提出,有效解决了机器学习模型训练过程中因试验数据缺失所引发的关键问题,并且其预测精度满足工程应用的需求。本研究为类似工程试验问题提供了宝贵的经验和方法论支持。本研究的主要成果及结论如下:
(1)通过引入微调网络层以及创新的损失函数,本研究成功克服了因试验数据缺失导致的机器学习模型训练中的核心难题,显著提升了模型的适应性和鲁棒性。
(2)本研究提出的迁移学习策略在试验数据不完整的情况下,能够高效且准确地预测波浪载荷,尤其在垂向弯矩的预测上,其精度普遍超过90%,大幅提升了预测的准确性。
(3)相较于传统仿真技术及其他机器学习方法,本研究方法在处理数据缺失、预测精度、计算效率以及模型泛化能力方面展现出显著的优势。
  • 国防科工局稳定支持基金资助项目(WDZC70202030201)
参考文献 引证文献
排序方式:
[1]
焦甲龙,陈超核,任慧龙. 真实海况下船舶水弹性响应及砰击载荷的大尺度模型试验研究[J]. 船舶力学2021, 2:138-145.
Jiao J L, Chen C H, Ren H L. Investigation of hydroelastic response and slamming loads ofships in real waves by large-scale model measurement[J]. Journal of Ship Mechanics, 2021, 2:138-145. (in Chinese)
[2]
陈帅,蒋彩霞,王子渊,. 基于迁移学习的多源波浪载荷融合方法研究[J]. 船舶力学2023, 10:1432-1444.
Chen S, Jiang C X, Wang Z Y, et al. Multi-source wave load fusion methodbased on transfer learning[J]. Journal of Ship Mechanics, 2023, 10:1432-1444. (in Chinese)
[3]
刘一夫. 超大型集装箱船波浪载荷大尺度模型试验研究[D]. 哈尔滨:哈尔滨工程大学,2019.
Liu Y F. Large scale model experimental study on wave load ofultra-large container ships[D]. Harbin: Harbin Engineering University, 2019. (in Chinese)
[4]
肖武,王海洋. 某大尺度新船型波浪载荷理论预报[J]. 船舶2017, 28(4):39-46.
Xiao W, Wang H Y. Wave loads prediction of large scale new type ship[J]. Ship & Boat, 2017, 28(4): 39-46. (in Chinese)
[5]
孙雪荣,汪战军,彭亚康. 小方型系数船舶的船体波浪载荷综合研究[J]. 舰船科学技术2017, 39(15):17-21.
Sun X R, Wang Z J, Peng Y K. The study of wave loads about small block coefficient ships[J]. Ship Science and Technology, 2017, 39(15): 17-21. (in Chinese)
[6]
任慧龙,陈亮亮. 大型水面舰船波浪设计载荷的模型试验研究[J]. 华中科技大学学报(自然科学版)2015, 43(8):79-82.
Ren H L, Chen L L. Model test investigation of design wave loads of large surface warship[J]. J. Huazhong Univ. of Sci. & Tech. (Natural Science Edition), 2015, 43(8): 79-82. (in Chinese)
[7]
徐言民,马国勤. 大幅波浪中船舶非线性运动与波浪载荷仿真[J]. 系统仿真学报2008, 9:2458-2461.
Xu Y M, Ma G Q. Simulation on nonlinear ship motions and wave loads with large amplitude waves[J]. Journal of System Simulation, 2008, 9:2458-2461. (in Chinese)
[8]
汪雪良,顾学康,祁恩荣,. 船舶波浪载荷预报方法和模型试验研究综述[J]. 舰船科学技术2008, 30(6):20-28.
Wang X L, Gu X K, Qi E R, et al. A state-of-the-art review on study of prediction methods and model test of wave-induced loads of vessels[J]. Ship Science and Technology, 2008, 30(6): 20-28. (in Chinese)
[9]
徐能兹,沈进威,陈瑞章,. S-175集装箱船舶波浪载荷的船模试验与线性切片理论计算的比较[J]. 中国造船1984(2):69-82.
Xu N Z, Shen J W, Chen R Z, et al. Model tests of s-175 containership on wave loads and comparisons between experiments and calculations of linear strip theory[J]. Shipbuilding of China, 1984(2): 69-82. (in Chinese)
[10]
Jiang C X, Liu Y B, Wang Z Y, et al. A multi-fidelity prediction model for vertical bending moment and total longitudinal stress of a ship based on composite neural network[J]. Journal of Hydrodynamics, 2023, 35:117-140.
[11]
Paik J K, Lee D H, Kim S J, et al. A new method for determining the design values of wave-induced hull girder loads acting on ships[J]. Ships and Offshore Structures, 2019, 14(Sup1):63-90.
[12]
Pena B, Huang L. Wave-gan: A deep learning approach for the prediction of nonlinear regular wave loads and run-up on a fixed cylinder[J]. Coastal Engineering, 2021, 167:103-902.
[13]
Kumar N K, Savitha R, Al Mamun A. Ocean wave characteristics prediction and its load estimation on marine structures: A transfer learning approach[J]. Marine Structures, 2018, 61:202-219.
2025年第29卷第10期
PDF下载
89
40
引用本文
BibTeX
文章信息
doi: 10.3969/j.issn.1007-7294.2025.10.002
  • 接收时间:2025-03-27
  • 首发时间:2026-03-26
  • 出版时间:2025-10-20
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2025-03-27
基金
国防科工局稳定支持基金资助项目(WDZC70202030201)
作者信息
    1.中国船舶科学研究中心,江苏 无锡 214082
    2.深海技术科学太湖实验室,江苏 无锡 214082
    3.西北工业大学,西安 710072

通讯作者:

通讯作者,E-mail:
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/cblx/CN/10.3969/j.issn.1007-7294.2025.10.002
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏